Table Saw Hazard Study
on Finger Injuries Due
to Blade Contact
JANUARY 2014
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Table of Contents
Executive Summary............................................................................................... 3
Introduction............................................................................................................. 6
Hazard Analysis....................................................................................................... 8
Injury Classification ............................................................................................. 13
Table Saw Injury Database.................................................................................19
Approach Speed Experiments......................................................................... 22
Disclaimer............................................................................................................... 32
Acknowledgments ............................................................................................. 32
Works Cited . ......................................................................................................... 33
Appendix A ............................................................................................................34
Appendix B ............................................................................................................36
page 2
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Table Saw Hazard Study on
Finger Injuries Due to Blade Contact
Executive Summary
The intent of this research was to understand the circumstances that may lead to hand/
finger contact injuries for operators of table saws and help identify critical parameters
that would define the hazard level. For the specific hazard of hand/finger coming into
contact with a spinning saw blade, two critical parameters were identified: first, hand
approach velocity, as an appropriate metric for the movement of the hand/finger
toward the spinning blade; and second, a medically guided estimate of maximum
cut depth of the finger that would help distinguish between simple and complex
lacerations and provide a positive impact on reducing the severity of hand/finger
injuries. Based on the research described within this report and the intent of reducing
hand/finger blade contact injuries for table saw operators, the following relationship
between approach speed and depth of cut (Figure 1) is recommended:
Figure 1: Recommended approach speed versus depth-of-cut relationship
depth of cut, mm
12
10
8
6
4
2
0
0
0.5
1
1.5
approach speed, m/s
page 3
2
2.5
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
• Finger contact with a table saw should not result in a depth of cut of more than 4 mm
(0.157 inches) at any approach speeds equal to or less than 1 m/s (3.28 ft/s).
• The table saw should be tested at a lower approach speed to ensure that the
performance of the safety technology does not degrade. This lower approach speed
can be based on the operational details of the safety technology or else set to a value
of 0.1 m/s (0.33 ft/s).
• In addition, the depth of cut should remain proportional to approach speeds beyond
1 m/s (3.28 ft/s). An upper approach speed limit of 2.5 m/s (8.2 ft/s) is recommended.
Therefore, at this approach speed, the acceptable depth of cut should be 12 mm (0.47
inches) or less.1
Note: The approach speed described above is the linear speed of a surrogate finger that is
traveling parallel to the table saw surface toward the midpoint of the exposed portion of
a blade set to a standard height.
Note: The depth of cut should be measured on a surrogate finger. The surrogate finger
must possess the key properties that trigger the safety device and allow for repeatable
depth-of-cut measurements.
The research described within this report provides the rationale for these
recommendations and is summarized next.
• For the depth of cut, it was necessary to find a quantitative threshold that might serve
as a reasonable boundary between simple and complex lacerations. This prescribed
value was based on measurements of key anatomical features within the human
finger and an understanding of the surgical and treatment implications as the depth
of cut increases. Since the focus of the finger injury research was on depth of cut,
specifics of blade width, tooth profile and other factors were ignored. However, depth
of cut is a primary factor in determining the injury severity, and reducing the depth
of cut will still help tremendously in addressing blade contact injuries regardless of
these other factors.
• Based on the approach speed estimates derived from the SawStop customer injury
database, the set point of 1 m/s (3.28 ft/s) was selected as typical of cases where the
injury might be considered a simple laceration. Though this database covers only
cabinet and contractor saws, all table saws use the same safety technology. Some
of the entries are based on self-reporting by users, yet the database tries to capture
a wide array of details about an incident involving blade contact injury. Moreover,
1 Includes a 20% buffer above the proportional value.
page 4
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
the database is unique by virtue of the data extracted from an electronic cartridge
that is part of the safety system, and so the SawStop database is the key source for
the approach-speed set point recommendation of 1 m/s (3.28 ft/s). However, extended
analysis of the database for more serious treatments,2 limitations of the database,
the experiments conducted in this research simulating reach-over and the incident
narratives suggest that injuries could also be occurring at higher speeds. For example,
one study found that amateur status was a significant predictor of tool-related injuries
for woodworkers (Becker et al., 1996). Incident narratives where a person "noticed a
wood panel falling off the edge of his table’ and then ‘lunged to catch it’ when ‘his
right forearm got caught on the blade" suggest higher approach speeds than walking
speeds for the person coming into contact with the blade.3 Another parallel exists
with OSHA4 standards covering mechanical power presses (29 CFR-1910.217). Though
the standard prescribed a hand movement of 1.6 m/s (63 in/s), research showed that
there were groups of users that exceeded these movements at speeds more than twice
that stated in the standard (NIOSH, 1987). Based on the simple experiments of this
research on reach-over, approach speeds in the range of 2.5-3 m/s (8.2-9.8 ft/s) may be
possible in the general population of table saw users, and any system that can mitigate
hand/finger blade contact injuries at speeds greater than 1 m/s will most likely have
a substantive impact on injury rates. Also, workpieces can be propelled at very high
speeds during kickback; however, the challenges with operator safety prevented actual
hand approach-speed measurements during this research program. Thus, in the area
of kickback, the effect of hand approach speeds still remains unresolved.5 However, to
impose a 4 mm (0.157 inches) depth-of-cut limit for speeds over 1 m/s (3.28 ft/s) may
create an undue constraint on safety technologies for table saws, and so instead a
speed-dependent cut threshold is recommended.
It is recognized that this research provides a reasonable, first-order estimate for two
critical parameters addressing the specific hazard of hand/finger blade contact with table
saws. As such, further research should be encouraged to build on this work and to improve
understanding of the circumstances in which table saw injuries occur, to the end that
significant findings can be compared against the analysis presented in this work and fill
in the gaps that advance public discussions, product development, policy decisions and
safety standards.
2 The SawStop manual states that "it is possible to be seriously injured even with the SawStop system.
3 Emotionally and financially, “it’s devastating.” – Adam Thull, in an article by Lily Fowler on Fairwarning.org, May 16, 2013.
4 US Occupational Safety and Health Administration.
5 Of the 1,316 entries in the SawStop customer injury database (provided in 2011), 172 were labeled as yes in the kickback question.
page 5
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Introduction
The US Consumer Product Safety Commission (CPSC) maintains a national database
(NEISS: National Electronic Injury Surveillance System6) of patient information collected
from participating 24-hour hospital emergency rooms to help investigate consumer
product-related injuries. In 2009, the CPSC conducted a special follow-up study of the
NEISS database focused on stationary table saw-related injuries treated in 2007 and
2008 (Chowdhury & Paul, 2011). The study involved an analysis of the NEISS data on
table saw injuries along with follow-up phone interviews to gather information about
the characteristics of the saws and other factors related to the injury incident.7 The study
estimated that 79,500 US hospital-treated injuries related to table/bench saws occurred in
2007-2008. Further breakdown of the data and survey results indicated that in 95.7% of all
cases, the operator of the saw was injured. In 88% of the cases where the operator of the
saw was injured, the injury was incurred from contact with a blade. Because in 94.5% of
all cases the motor was running, and 91.2% of all cases involved cutting of a wooden board,
it can be assumed that most injuries cited are due to contact with a rotating blade while
cutting wood. In 89.1% of all cases, the finger was the injured body part, followed by the
hand in 6.8% of all cases. The injuries were classified as lacerations in 64.8%, fractures in
12.2% and amputations in 10.5% of all cases.
In a search of other publicly available literature8 on the topic of table saw injuries9, 10, 11 a
paper was found that also analyzed data from NEISS over a time period extending from
1990-2007 and covering adults and children (Shields et al., 2010). This paper basically
reached a similar conclusion: most table saw-related injuries were a result of hand/finger/
thumb contact with the saw blade. One study (McGonegal, 2012)12 estimates that the number of
table saw injuries during 2007-2008 was lower than that provided by the CPSC study.
6 http://www.cpsc.gov/cpscpub/pubs/3002.html
7 See Appendix B for sample narratives.
8 Literature reviews based on Google search engine using English-language sources.
9 http://www.hospital-data.com/accidents/841-bench-or-table-saws/index.html#b is one website that claims to graphically display data on bench or table
saw accidents taken from NEISS.
10 http://www.sawaccidents.com/p1.htm is a website claiming to collect data from visitors on table saw accidents.
11 SawStop, a table saw manufacturer, collects data from its customers who experience blade contact accidents (over 1,300 incidents) and has provided this data
to the CPSC as part of its response to an ANPR on table saw blade contact injuries; http://www.regulations.gov/#!documentDetail;D=CPSC-2011-0074-1106.
12This document was provided by Power Tool Institute and can be accessed at http://www.regulations.gov/#!documentDetail;D=CPSC-2011-0074-1081.
page 6
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Looking at the total number of table saws in the US, the CPSC – based on information
provided by the Power Tool Institute – estimates that there were about 9.5 million units
in 2007/2008 with lifetimes ranging from six years (consumer-grade bench saw) to 17
years (contractor saw) to 24 years (cabinet saw) (CPSC, 2011). As an estimate on the
economic impact of table saw injuries, the CPSC (CPSC, 2011) and others (Hoxie et al.,
2009; McGonegal, 2012) have calculated societal costs in the range of $30,000 to
$40,000 per incident.13
Currently, table saws in the US market are certified in accordance with the Standard for
Stationary and Fixed Electric Tools, UL 987. This ANSI-approved, voluntary, consensus-based
safety standard has been updated several times since its inception in 1971, with new
requirements for safety features such as guards and riving knives, and was most recently
updated in 2009. However, the standard does not presently require table saws to
incorporate safety devices that might mitigate injuries once a hand/finger is in contact
with a rotating blade.
In September 2011, UL started a research project and formed a Working Group (WG) by
inviting a group of outside experts on table saw operation, table saw safety and table
saw injuries. The scope of the research project was to develop a set of performance
requirements for active safety technologies through a rational process to help update
safety standards for table saws.
These performance requirements would define the conditions covering a specific hazardous
condition and would serve as the prerequisites for any new test protocols for the table saw
safety standard. In this research, focusing on the specific hazard of a hand/finger coming
into contact with a moving saw blade, the performance requirements would be based
on two building blocks: one would be an understanding of the circumstances in which
the hand/finger of the operator could come into contact with the blade and use a single
parameter, hand approach velocity, as an appropriate metric; the other would require
some medically guided estimate of maximum cut depth of the hand/finger that would
help distinguish between simple and complex lacerations and provide a positive impact on
reducing the severity of hand/finger injuries.
13 This report does not attempt to critique or evaluate any analyses conducted on injury rates due to table saws or the economic impact of such injury rates. The starting
point of this research is simply that there are well-documented hazards involving table saws whereby the hand/finger of the operator can come into contact with a
rotating blade.
page 7
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Hazard Analysis
To properly investigate a hazard, it is
that generates harm (injury) to the
For hazard analysis of the table saw,
grounded on established engineering
the hazardous source (rotating blade).
selected. The FTA is a deductive process
important to follow a methodology
and risk analysis principles. For this
susceptible part (hand/finger) from
Though it is not being addressed in the
current work, the diagram demonstrates
reason, the first step of the project
involved using some specialized
tools and techniques to analyze and
document harm from hazards and
protections against hazards within a
well-defined scope. Figure 2 displays
the general framework for any hazard
whereby energy is transferred to a
how the first and most effective step
is prevention of energy transfer or
elimination of the hazardous source.
Another possible strategy is guarding,
preventing the transfer mechanism
from acting on the susceptible part. For
the hazard of a hand/finger coming in
contact with a moving blade, reducing
susceptible body part, resulting in
harm. For example, when a hand is
approaching a moving blade, physical
contact is the transfer mechanism
the susceptibility is one means of
reducing the risk of serious harm.
the Fault Tree Analysis (FTA) tool was
starting with a high-level failure and
working through the sequence of events
to determine possible root causes
(Vesely, 2002). The approach relies on
a treelike graphical representation to
visually communicate the linkages
between the different layers of analysis.
The FTA captures the current state of
knowledge regarding the root causes
for a hazard and should be viewed as a
working document.
Figure 2: Hazard-based safety engineering conceptual diagram
ENERGY
HARM
HAZARDOUS
SOURCE
TRANSFER
MECHANISM
SUSCEPTIBLE
PART
ELIMINATE
GUARD
WARN
page 8
REDUCE SUSCEPTIBILITY
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Fault Tree Analysis
for Blade Contact Injury
In developing an FTA, it is necessary
help guide the subsequent FTA based
and with another solid line connects the
with the blade.
energy transfer during the cutting action.
on a scope covering hand/finger contact
to first identify the key functional
The main components of a table saw
the components that help define the
and blade, along with passive safety
components and interactions between
scope and boundaries of the analysis.
Figure 3 shows a detailed part breakdown
for a particular commercial table saw.
This level of detail is unnecessary for
the purposes of this research. However,
selectively simplifying and grouping the
components of a product to extract the
key features is an important first step. The
outcome of this process is captured in a
simplified functional block diagram (FBD),
Figure 4, for a generic table saw with the
key components and connections that will
Figure 3: Components of a commercial
wood sample to the blade, indicating
The riving knife is shown acting between
include a motor, table/base, power switch
features. Energy (mechanical or electrical)
is transferred among the components
represented by solid lines in the FBD. For
the wood sample and the blade (tip of
the object labeled "Riving Knife" in Figure
4). This represents the role of the riving
knife in trying to reduce the probability of
kickback during cutting of the work-piece.
example, the motor drives the rotation
The presence of a guard, required by UL
an operator would be pushing against
contact, mainly from the top, back and
of the blade and, when the saw is in use,
a work-piece along the table surface
987, affords some protection against
side of the blade. In the FBD, a solid
as it is being cut by the rotating blade.
The purpose of the riving knife, required
by UL 987, is to reduce the probability
of kickback.15 In the FBD, a solid line
connects the operator to the wood sample
line is shown directly connecting the
operator to the blade, with the guard
shown intervening in the middle (tip of
the arrow labeled "Guard" in Figure 4).
This graphic represents the role of the
Figure 4: Simplified functional block diagram for generic table saw
table saw 14
WOOD
SAMPLE
BLADE
RIVING
KNIFE
POWER/
SWITCH
OPERATOR
GUARD
MOTOR
TABLE/BASE
14 http://www.toolsandpartsdirect.co.uk/Elu-EMTS711-Type-1-Table-Saw-Spare-Parts__p-5574.aspx.
15 The definition of kickback, according to UL 987, is "sudden reaction to a pinched, bound or misaligned work-piece with respect to the saw blade, which causes the
work-piece to be propelled by the blade."
page 9
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
guard in passively shielding the operator
hand/finger is coming into contact with a
some descriptive summaries of numerous
circumstances. As noted in the CPSC study
of wood stock without the presence of a
for discussions by the WG on the possible
from contact injury with the blade in some
(Chowdhury & Paul, 2011), a guard was
present in 34.3% of the cases involving blade
contact injuries. This shows that, even with
a guard present, based on current designs,
blade contact injury is possible. However,
since most blade contact injury cases cited
in the CPSC study were with no guard
present, the no- guard scenario is considered
to be more comprehensive for the purposes
of this hazard analysis.
With the FBD completed, the scope for the
next step, developing the FTA, covers the
hazardous condition in which an operator’s
moving blade during a cutting operation
incidents and helped serve as background
guard. Details such as the type of cut, state
scenarios that could lead to a blade contact
of the blade and others were not deemed
injury. Of course, post-incident interviews
pertinent for this analysis.
and self-reporting surveys of people
involved in injurious events (such as those
One of the key challenges in trying to
understand table saw injuries is knowledge
of the diverse behavior of a human operator
at a home, work or other setting where
there are no independent records of the
circumstances and the measurements of
key parameters during the incident (like
that provided by flight data recorders in
airplanes). The CPSC survey (Chowdhury
& Paul, 2011) noted earlier did provide
conducted by the CPSC and SawStop),
can be valuable in a hazard analysis, but
they must be carefully analyzed and
sometimes require corrections.16 The
potential limitations of self-reporting
and the effect of recall duration for
interviews should be understood when
designing such surveys and analyzing any
subsequent data (Landen et al., 1995).
Figure 5: FTA of generic table saw with top-level hazard involving hand or finger blade contact injury
INJURY TO HAND/FINGER FROM
CONTACT WITH TEETH OF
MOVING BLADE
1
UNINTENDED CONTACT OF
HAND/FINGER WITH SAW BLADE
MOVING
BLADE
2
NO GUARD
NON-KICKBACK
CONDITION
NO GUARD
4
3
HAND CONTACTING
BLADE WHILE
PUSHING WORK-PIECE
WITHOUT SLIPPING
CONTACT WITH BLADE
DUE TO SLIPPAGE
WHILE PUSHING
WORK-PIECE
KICKBACK
CONDITION
CONTACT WITH BLADE
WHILE REACHING
OVER DURING
CUTTING ACTION
MOVEMENT OF
HAND/FINGER
RELATIVE
TO WORK-PIECE
OPERATOR
ACTION
OPERATOR AWARENESS,
REACTION TIME, ETC.
16 An analysis of data recorders from Prius vehicles involved in "sudden acceleration" accidents where drivers stated they had applied the brakes has shown no evidence of braking and suggests some faulty recall of actions during the inciden;: http://www.nhtsa.gov/PR/DOT-16-11.
page 10
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
The building of the FTA starts with a high- level hazard: an injury from a hand/finger coming
into contact with a moving blade (Figure 5). Below this top event is the first gate (an AND
gate labeled as ‘1’), which requires that all items feeding into it from the immediate lower
level of the fault tree occur for the high-level hazard to be realized. So, for a blade contact
hand/finger injury to occur, both "unintended contact of hand/finger with saw blade" AND
a "moving blade" are required. The ‘moving blade’ event is not developed further as it would
not provide further insight into the injuries for these purposes. Of course, injuries can still
occur with a stationary blade as noted in one of the incidents in Appendix B. Already, one
can see the challenges in trying to capture circumstances that cover a wide range of ways
in which the operator of a table saw could be injured through contact with a blade. One
can always imagine a circumstance where an operator might slam his or her hand/fingers
onto a stationary blade. The question that must continually be asked at each stage of the
FTA is whether a circumstance would help inform the hazard analysis and not put undue
requirements on safety systems for table saws. However, the other event under the AND
gate 1, "unintended contact of hand/finger with saw blade", needs to be developed further
to help capture the circumstances leading to the specified hazard.
The gate labeled 2 is an OR gate, and indicates that any one of the immediate lower events
leading into it are sufficient for the event above it to occur. Therefore, the "unintended
contact of hand/finger with saw blade" event could have one of three immediate causes.
The first event, "operator action," leading into this gate could simply mean that the blade
is rotating and, without any cutting of stock, the operator inadvertently brings his hand/
finger in contact with the blade. Other researchers have looked into this issue and have cited
reasons such as "being in a hurry, not realizing the hand was in a hazardous area, misjudging
time and distance to avoid an injury, shifting work materials or tools and attention not fully
on task" (Sorock et al., 2001) that would fall under this general event. This item will not be
developed further as the analysis by the CPSC summarized in the introduction indicated that
most injuries occur during cutting operations. It is also likely that some aspects of operator
action will show up in other branches as the FTA is further developed.
The two other events, kickback and non-kickback conditions, describe two general categories
that cover work-piece cutting operations when the hand/finger of the operator could come
into contact with the blade. The choice of these categories was an attempt to develop a
more tractable analysis. To list every conceivable way in which an operator could be coming
into contact with a blade would create a very complex tree without necessarily improving
the analysis outcome. One revelation when discussing the scenarios with the WG was the
difficulty in describing a "representative" circumstance for blade contact injuries with high
confidence. But this is not really surprising, considering how the state of the operator plays
a large role in these incidents.
page 11
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
For the kickback and non-kickback events, there is an adjoining oval with the label "No
Guard." This oval indicates that the event occurs under the conditions described in the oval.
This simply reflects the scope of the analysis described earlier in this section. The kickback
and non-kickback events are considered to be taking place without the presence of a guard.17
For the kickback condition, both "movement of hand/finger relative to the work-piece"18 that
is being kicked back and (Gate 4) the "operator awareness of and reaction" to the sudden
change in movement of the work-piece determine how fast the hand is moving and whether
it contacts the blade. These events under the kickback condition are not developed further
for reasons to be discussed in the Approach Speed Experiments section.
For the non-kickback condition, there are basically three general circumstances in which the
hand/finger could contact the saw blade. First condition: an operator is pushing a work-piece
into the blade during a cutting process, and it is possible that the operator is unaware of his/
her hand moving very close to the blade until there is contact. Second condition: an operator
is pushing a work-piece into the blade during a cutting process, and it is possible that the
hand will slip and move toward the saw blade at a speed faster than the work-piece speed.
Third condition: during a cutting operation, the operator is reaching over the saw blade and
being distracted in the process so that a hand/finger comes into contact with the rotating
saw blade. Notice in all conditions that operator awareness and reaction influence the
motion of the hand/fingers.
With an initial FTA completed, and having identified four different general circumstances
(highlighted in red in Figure 5) in which an operator of a table saw may experience an injury
of the hand/fingers due to blade contact, the next step is to attempt to simulate these
circumstances and record/analyze the needed measurements that would help provide
quantitative estimates of the key metric, hand approach velocity. Standard EN 999:1998,
Safety of Machinery,19 which cites a value of 1,600-2,000 mm/s as the typical walking speed
and hand/arm approach speed, is used to determine distance thresholds for "sensing or
actuating devices of protective equipment in a danger zone". Yet, it was not known how
well walking speed might represent hand movement during wood-cutting operations. So a
series of experiments was designed and run to simulate some of the conditions identified in
the FTA. In addition, a customer incident database for table saws by one manufacturer was
analyzed. This database was provided by SawStop LLC, which sells a table saw system that
"detects when someone accidentally contacts the spinning blade and then stops the blade
in milliseconds."20 Both the experiments and the injury database were examined in detail as
they relate to approach velocity for hand/fingers. However, the next section first covers the
classification of injuries and how it informs a possible injury threshold.
17 It is still possible for injuries to occur even with the presence of the passive safety devices.
18 This can include the case where the relative movement is zero; that is, the hand is moving with the work-piece without slippage.
19 This standard is now superseded by EN ISO 13855.
20 http://www.sawstop.com/wp-content/uploads/sawstop_whitepaper.pdf.
page 12
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Injury Classification21
Table saw injuries can involve lacerations
simple lacerations generally heal
to their distal location (Figure 6) and
lacerations associated with injury to the
of varying depth in fingers, and owing
small size, fingers are especially prone to
injury. Finger lacerations can be grouped
according to structure damage: simple
lacerations or complex lacerations,
which include tendon, nerve or vascular
involvement and amputation (Figure
7). Simple lacerations that damage only
the epidermis and dermis can usually
be treated at home with simple wound
care, but may require emergency
department attention. In either case,
Figure 6: Phalanges of the human finger
uneventfully. The exception is fingertip
fingernail bed. Careful reconstruction
must be undertaken to avoid fingernail
bed scarring that can lead to painful or
deformed finger nail regrowth. Deeper
lacerations along the length of the finger
disrupt more-vital structures. Severed
tendons, nerves and, vessels require
skilled microsurgery to ensure optimal
functional and aesthetic results. Surgery
may require a hospital stay and lengthy
occupational therapy.
Figure 7: Internal structure of the finger
21 This section is extracted with some modifications from a full report: "Table Saw Injuries: epidemiology and a proposal for preventive measures," by Dr. Kevin Chung and
Melissa Shauver, Department of Surgery, University of Michigan. This work was commissioned and funded by UL. Furthermore, this work has been accepted for publication
in a peer-reviewed medical journal, Plastic and Reconstructive Surgery.
page 13
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Review of Medical Literature
eliminated, leading to a much-shortened
Using CINAHL and PubMed databases to
list of citations (Figure 8).
search the English-language literature for
One study (Al-Qattan, 2012) followed
saw” in the title or abstract resulted in the
mean of 16 weeks (range: 14-25 weeks).
articles with the keywords “saw” or “table
following: after filtering the searches and
removing duplicates, the list comprised
64 citations consisting of articles about
table saw injuries, specifically to the upper
extremities. If an article included several
types of saws or tools, the data regarding
table saws had to be extractable. Case
reports or articles about repair technique
after table saw injuries were also
16 patients who had saw injuries for a
The patients’ injuries could be described
joint. In the first group, final range of
motion was 54% of that of the normal
finger (range: 29%-78%). In the more
severely injured second group, range of
motion was 32% of the normal finger
(range: 14%-57%).
in one of two injury groups caused by a
A 2002 nationwide study found that 20%
neck fractures with concurrent extensor
attributed to table saws, more than any
saw. The first group sustained phalangeal
tendon injury. The second group also
sustained fractures and tendon injury, but
additionally sustained nerve transection.
None of the patients recovered full range
of motion at the distal interphalangeal
of nonoccupational amputation could be
other consumer product, anything sold
for the personal use or consumption of
consumers in a nonoccupational setting
(Conn et al., 2005). A nationwide survey
dispersed in a woodworking magazine
Figure 8: Literature search and screening process
PUBMED
SAW [TITLE]
N=997
PUBMED
TABLE SAW [TITLE/ABSTRACT]
N=8
CINAHL
SAW [TITLE]
N=287
CINAHL
TABLE SAW [TITLE or ABSTRACT]
N=287
PUBMED
(SAW [TITLE] or TABLE
SAW [TITLE/ABSTRACT]
N=1,003
CINAHL
(SAW [TITLE] or TABLE
SAW [TITLE or ABSTRACT]
N=289
PUBMED
and TRAUMA [MeSH TERM]
N=63
CINAHL
and TRAUMA [MeSH TERM]
N=4
N=64
ANALYSIS COHORT
N=9
REJECTED
MEDICAL SAWS
N=18
OTHER SAWS
N=29
CASE REPORTS/ TECHNIQUE
N=5
OTHER USES OF “SAW”
N=3
page 14
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
found that 42% of reported injuries
The third and most notable group injured
department (ED) (Knight et al., 2000).
the cause of the highest proportion of
woodshop courses. In a longitudinal
requiring ED treatment. This is a higher
were caused by a table saw, and it was
amputations (39%) (Justis, 1987). A survey
of Albuquerque-area professional and
amateur woodworkers reported that
the table saw was involved in for 31%
of injuries and was responsible for the
highest proportion of injuries requiring
medical attention (21%) (Becker et al.,
1996). Table saw injuries resulted in
hospitalization 7% of the time; the mean
for all consumer products is 4%
(CPSC, 2011).
Individuals injured while using a table
saw can be grouped into three broad
categories. The first two are occupational
and nonoccupational injuries. Both of
these groups are overwhelmingly male.
The percentage of injuries sustained at
work ranges from 31% to 46% (Becker
et al., 1996; Justis, 1987; Waller, 1990).
Injured workers had a mean age of 40
years. Hobbyist or amateur woodworkers
tend to be older than professionals
(mean age approximately 50 years). A
Poisson regression controlling for age,
experience, type of woodworking activity
and training found that amateur status
was a significant predictor of injury
(p<0.0001) (Becker et al., 1996). Despite
this, work-related and non-work-related
injuries have a similar injury pattern,
recovery course, and costs of care (Hoxie
et al., 2009).
by table saws is minors injured in school
sample of nonoccupational table saw
injuries, only 3% of the samples were
minors, but more than half of them were
injured at school (Shields et al., 2010).
This is especially interesting in light of
the fact that per US Department of Labor
regulations, minors are not permitted
to use power-driven woodworking tools
or saws, metal-forming machinery, or
punching machines in the workplace.22
Yet this equipment is regularly used
by children as young as 11 years old in
middle-school and high-school shop and
industrial arts classes. A Utah statewide
examination of injuries at schools found
that 7% of injuries occurred in shop and
that 31% of injuries were caused by saws
(table saw and band saw 12% each and
7% by other saws) (Knight et al., 2000).
Improper equipment use was cited in 38%
of cases. Injured students were primarily
in grades 8 and 9 (42%) and were mostly
male (87%). The average time missed
from school was one-half day (range:
0-36). Fingers and thumbs were the
most frequently injured body parts (64%)
followed by hands/wrists (13%) and eyes
(6%). Lacerations were the most common
injury (71%). Burns (6%) and abrasions
(5%) were also experienced (Knight et al.,
2000). Twenty-seven percent of injured
Table saws were involved in 15% of injuries
proportion than the proportion of injuries
caused by table saws, indicating that table
saws cause more serious injuries than do
other types of equipment. Seven students
were admitted to the hospital following
a shop class injury. Six of those students
were using table saws, and in four of the
six cases, improper use of the table saw
was cited as the cause of the injury. Two
patients sustained a finger amputation
and one patient sustained a thumb
amputation; the other three patients
sustained hand or finger lacerations with
tendon, nerve and/or bone involvement.
Another study (Beavis et al., 2006)
specifically looked at hand injuries that
occurred in shop class. All the patients
were male, with a mean age of 16 years
(range: 12-18 years). Sixty percent of
the injuries were caused by table saws,
resulting in two index finger amputations,
one index and long finger amputation,
and lacerations and/or abrasions of
various degrees of severity. The outcome
of treatment was assessed at an average
of 22 weeks (range: 3-66 weeks) and
revealed sensory and range of motion
deficits in patients who had tendon, nerve
or artery repair and sensitivity in patients
treated with revision amputation.
students were treated at an emergency
22 Child Labor Provisions for Nonagricultural Occupations Under the Fair Labor Standards Act in: US Department of Labor Wages and Hours Division, ed.; 2010.
page 15
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Measurements on Finger Structure
Because the depth of finger laceration has a direct effect on the structures injured, it
is important to understand the depth of key anatomic structures relative to the skin’s
surface. In an extensive review of the literature, no sources were found providing anatomic
measurements quantifying the relationship of vital structures in the fingers to the skin.
Therefore, results were generated for this research by examining magnetic resonance
imaging (MRI) scans from five patients from the Department of Surgery at the University
of Michigan hand surgery practice who had undergone MRI scans for a hand condition
unrelated to the fingers. Patients were selected for their similarity to the average sufferer of
a table saw-related hand injury: middle-aged males (Becker et al., 1996; Shields et al., 2010;
CPSC, 2011). Patients were male, with a mean age of 54 years (range: 42-74). MRI was of the
dominant hand in two cases and of the nondominant hand in one case. In two cases hand
dominance was not known. Using PACS radiology software, the distance from the surface
of the skin to the neurovascular bundle, flexor tendon and bone was measured for each
digit from both the radial and ulnar side. Measurements were taken at the midpoint of each
phalanx (Figure 9) and the mean and standard deviation were calculated.
Figure 9: Sample MRI image of human fingers with anatomical structures labeled on the index finger
on the index finger
page 16
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Generally, structures were deepest at the proximal phalanx of the long finger and
most shallow at the distal phalanx of the little finger. The neurovascular bundle, which
contains the nerves and arteries, is the structure closest to the skin’s surface (Table 1).
Its depth ranged from 4.3 mm at the distal phalanx of the little finger to 7.1 mm at the
proximal phalanx of the index finger. Tendon depth ranged from 4.6 mm to 7.1 mm
(Table 2), whereas bone was located at 6.3 mm to 13.5 mm (Table 3) in depth. Figure 10
shows graphically the meaning of these measurements based on data from one patient.
Table 1: Mean (+/- standard deviation) distance from skin’s surface to neurovascular bundle
measured at the phalanx midpoint (mm)
Proximal
Middle
Distal
radial
ulnar
radial
ulnar
radial
ulnar
Thumb
4.8 ± 0.6
5.1 ± 0.8
--
--
4.4 ± 0.3
5.8 ± 0.7
Index
7.1 ± 0.5
6.2 ± 1.1
6.2 ± 0.3
5.4 ± 0.3
4.6 ± 0.7
4.3 ± 0.9
Long
6.5 ± 0.7
6.5 ± 0.6
5.8 ± 0.7
6.1 ± 0.4
4.9 ± 0.4
4.4 ± 0.7
Ring
6.5 ± 0.6
5.1 ± 0.8
4.9 ± 0.5
4.8 ± 0.4
4.6 ± 0.3
4.8 ± 1.2
Little
4.6 ± 0.9
4.8 ± 0.6
4.5 ± 0.3
4.4 ± 0.6
4.7 ± 0.6
4.3 ± 0.6
Table 2: Mean (+/- standard deviation) distance from skin’s surface to flexor tendon measured
at phalanx midpoint (mm)
Proximal
Middle
Distal
Thumb
5.1 ± 1.0
--
6.7 ± 1.6
Index
7.1 ± 1.3
5.8 ± 0.5
5.2 ± 1.4
Long
7.1 ± 1.4
6.9 ± 0.6
6.3 ± 0.5
Ring
6.5 ± 0.7
6.0 ± 1.1
5.4 ± 0.9
Little
5.1 ± 0.7
4.6 ± 0.6
4.7 ± 0.4
Table 3: Mean (+/- standard deviation) distance from skin’s surface to bone measured
at the phalanx midpoint (mm)
page 17
Proximal
Middle
Distal
Thumb
11.3 ± 1.7
--
8.8 ± 2.9
Index
13.2 ± 1.3
10.8 ± 0.6
6.6 ± 1.5
Long
13.5 ± 0.7
11.9 ± 1.0
9.6 ± 0.7
Ring
12.3 ± 1.2
10.7 ± 0.8
8.6 ± 0.8
Little
10.0 ± 1.0
7.7 ± 1.6
6.3 ± 0.7
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Because it is virtually impossible to make contact with a rapidly rotating saw blade
and avoid any injury at all, a simple finger laceration may be an acceptable goal for
a saw injury; these injuries can be managed in the ED with little expertise or may
require only simple home wound care because these cuts generally heal uneventfully.
A deeper cut will need surgery to repair the structures damaged, requiring increasing
levels of skill at increasing cost. Based on the measurements presented in this research,
a depth of 4 mm (0.16 inches) is the maximum depth for a cut to a finger before serious
injury is sustained.
Figure 10: Cross section of the human index finger with anatomical structures labeled.
Distances are from the skin’s surface at the midpoint of the distal phalanx for a
single patient (mean +/- standard deviation).
page 18
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Table Saw Injury Database
One table saw manufacturer, SawStop,
has been selling table saws equipped
with a capacitance-based sensing system
for many years (Figure 11). The SawStop
its entries. Instead, the database is used
to estimate the hand approach speed,
making clear where data is extracted
directly from the database and where
assumptions are made.
In the Injury Classification section, a 4
mm depth of cut was recommended as a
possible threshold value distinguishing
between simple and complex lacerations.
So only data where the treatment
product line includes contractor saws and
Analysis: To use the database to estimate
description might suggest a simple
but no bench table saws. During this time,
are required: depth of cut and "total time"
is analyzed and a 4 mm depth of cut
out a survey and return the safety system
when the hand first contacts the blade
entries, 124 had unspecified treatment.
incident triggering the safety system. The
the hazardous condition. For the depth
of the triggering event. Information in the
any quantitative information but instead
data (extracted from the electronic
as a starting point, database entries are
The database in this study included 1,316
of the treatment type. SawStop uses
industrial and professional cabinet saws
hand approach speeds, two sets of data
laceration (items 1-8 in the list above)
SawStop has encouraged customers to fill
(TT). The TT is defined as the time from
is assumed for each case. Of the 1,316
cartridge whenever they experience an
to when the system is able to remove
cartridge records the electrical signature
of cut, the database does not provide
database is a combination of processed
provides some general categories. So,
cartridge) and customer responses.23
first filtered according to the description
incidents since 2005.24
the following categories to describe the
The database contains information about
table saw incidents based on a survey
filled out and returned by SawStop
customers. However, as noted previously,
treatment required for the injury incurred
from contacting the saw blade and
activating the safety mechanism:
1. None
post-incident surveys can suffer from
2. Cleaned
analyzed in detail to assess the validity of
3. Antiseptic
errors, and the database has not been
Figure 11: Illustration of SawStop safety device
4. Band-Aid
5. Bandage
6. Bandaged at hospital or clinic
7. Stitches
8. In-house/first aid
9. Doctor visit
10.Doctor amputated finger tip
11. Not specified
Next consider how to calculate TT. Again,
the SawStop database does not directly
provide all the information needed for TT,
but does provide one of its components.
This component, called "time between
contact and detection" in the database,
represents the time interval between
initial changes to the signal that are likely
due to first contact (or soon after) of the
hand/finger (or any object) with the blade
and when contact is finally confirmed by
the SawStop algorithm. This value is based
on processing of the cartridge data. Once
object/blade contact is established by
the SawStop algorithm, then the system
activates a metal brake that stops the
blade from rotating. Therefore, TT is the
sum of the "time between contact and
detection" and the time it takes to stop
the blade from presenting a hazard.25
Continuing on, the data column labeled
"time between contact and detection"
is examined and another filter is
applied to remove the empty entries.
The final outcome is shown in Figure
12, a histogram of the "time between
23 A copy of the header information from the database is included in Appendix A.
24 These numbers are based on the number of entries in the database provided to the WG by Steve Gass (SawStop LLC) in October 2011.
25 The total time is affected by approach speed up to the limits of the electronics.
page 19
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
contact and detection" for the category
defined by a 4mm depth of cut and that
To develop the above analysis a bit
of treatment that might be considered a
the "total time" for this system to remove
further, consider treatment categories
simple laceration. Most of the time data
the hazard is (1 ms + 3 ms) 4 ms, then the
that might suggest a cut deeper than 4
is 10 ms or less. Assuming a fixed depth
estimated hand approach speed would
mm. This would include categories 6, 7, 9
of cut, a lower value of TT will lead to a
be 1 m/s (3.3 ft/s) for incidents involving
and 10. In these cases (69 entries), it may
more conservative (higher) estimate of
SawStop customer injuries for the
be possible that the cut extended up to
approach velocity. Examining the data
selected treatment categories.
the neurovascular bundle. Referring to
in more detail, Figure 13, it appears that
Since there is only a single value for the
1 ms could be a reasonable conservative
braking time, and it was not from the
threshold time that would cover most of
electronic cartridge, it is important to
these cases.
consider how the hand approach speed
Now to estimate TT, the additional time
might change based on a variation in the
for the brake mechanism to activate
braking time. If a lower limit of 2 ms for
and stop the blade is needed. This time
the braking time (as increased braking
interval is not available in the database,
time only results in lower approach
and only a single value of 3 ms has been
speeds) and system linearity are assumed,
provided by SawStop.26 Assuming that all
then the estimated approach speed would
these cases involve a simple laceration
be 1.3 m/s (4.3 ft/s).
Table 1, the possible depth of cut without
damaging this bundle could be as high
as 7 mm for the index finger, a finger that
is typically injured. Figure 14 shows the
time data for these selected treatment
categories, and once again 1 ms is a fair
estimate for "time between contact and
detection". Assuming a "total time" of 4
ms and a depth of cut between 6-7 mm
leads to estimated approach speeds in the
range of 1.5-1.75 m/s (4.9-5.7 ft/s).27
Figure 12: Histogram of contact-to-detection times of filtered SawStop customer injury database
CONTACT-TO-DETECTION
800
700
frequency
600
500
400
300
200
100
0
0 20
80
160
240
time (ms)
320
400
26 The SawStop owner’s manual (10” Professional Cabinet Saw Model PCS31230) states that "an improperly positioned brake could increase the time required to stop the
blade in the event of accidental contact." The same manual also states that the "blade will be stopped in about 3-5 milliseconds (coarse-toothed blades stop more quickly
than fine-toothed blades such as plywood blades)."
27 There was a single case labeled "doctor amputated finger" with a "time between contact and detection" of 0.2 ms. The amputated finger was the left ring finger.
page 20
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Figure 13: Close-up of data in Figure 12
Discussion
CONTACT-TO-DETECTION
Before this section ends, some possible
300
caveats of the database are discussed.
Approach speeds presented in this section are
250
estimates, and not direct measurements, with
assumptions identified at each step. Another
frequency
200
possible limitation in extending these results
to cover the general table saw user population
is that the current SawStop product line does
150
not include bench table saws. 29 Most of the
incidents in the database occur at a "workplace
100
or school," since "most SawStop table saws are
sold to professionals, schools and government
50
0
entities."30 As noted before, one study found
that amateur status was a significant predictor
0 1
of injury (p<0.0001) (Becker et al., 1996). There
20
is the possibility that amateur woodworkers,
time (ms)
who (according to the CPSC report (CPSC, 2011))
are likely to include owners of bench table
Figure 14: Histogram of time between contact and detection for the treatment
saws, "may discover dangerous and difficult
categories 6, 7, 9 and 10
operations only by actually experiencing near
accidents or problems." This might mean an
25
increased probability of injury and/or higher
injury severity. The latter assumption would
imply that amateur table saw users could
20
contact the blade under higher approach
frequency
speeds than those of professional users, the
typical owners of cabinet and contractor saws.
15
Nevertheless, despite these remarks, the
database is a very helpful piece of the puzzle in
10
framing performance requirements. However,
this database does not preclude the need
for the hazard analysis and experiments to
5
simulate circumstances in which an operator
of a table saw may contact the blade, to
0
possibly help define reasonable upper limits
0
5
10
15
20
25
30
35
time between contact and detection (ms)
for approach speeds. The next section details
experiments designed to simulate some of
the circumstances captured in the FTA and
hopefully advance knowledge about table saw
safety and contribute to the discussions and
28 Analysis by SawStop of the CPSC survey results suggests that a majority of injuries occurred on bench table saws.
See Comments and Information Responsive to the ANPR for Table Saw Blade Contact Injuries by SawStop, LLC, US
CPSC, Docket No. CPSC-2011-0074, March 16, 2012.
29 Ibid.
page 21
updating of the table saw safety standards.
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Approach Speed Experiments
From the hazard analysis, four general
scenarios were identified in which a table
saw operator’s hand/finger may come
into contact with a moving blade; these
require further study:
1.
Slippage of the hand that is
pushing a work-piece toward
the rotating blade
2.
Hand that is pushing (without
slip) a work-piece toward the
rotating blade
circumference of the blade, and another
designed in an attempt to simulate the
the blade (radial). When the hand/finger
might provide a reasonable estimate of
component directed toward the center of
contacts the blade, it is mainly the radial
hand speeds during certain conditions.
drives the hand/finger deeper into the
is to help inform the understanding of
component of the approach velocity that
The intent of estimating approach speeds
blade and determines the injury level.
the conditions under which most injuries
One key challenge is that the simulation
must include the realities of human
cognitive awareness and reaction times. In
any incident reconstruction, the operator
reaction time is an important factor.
Therefore, human subjects must be used,
3.
Saw blade reach-over by the operator
but in a test where minimal instruction is
4.
Movement of hand toward the blade
react differently. Also the simulation must
during a kickback condition
In each of these scenarios, one key
determinant of the degree of hazard
posed to an operator’s hand/finger is
the approach velocity. Using the center
of the blade as a reference point, as
the hand/finger contacts the blade,
the approach velocity can be broken
down into two components (Figure
15): one component tangential to the
motion of the human hand in a way that
given that might prime the candidate to
not present a hazard to the test subject.
To ensure a safe test protocol in running
the experiments to simulate hazardous
conditions, the following was decided: for
the slippage and reach-over experiments,
no blade should be present, and for the
kickback experiment, no human subject
should be in contact with or in proximity
of the work-piece. Within these safety
might occur and, in turn, contribute
toward the development of performance
requirements. However, for practical
reasons, it was difficult to run a full
demographic study of the population of
table saw users. Instead, a small sample
of table saw users was selected based
on access and availability.30 As a final
economy in the test plan, scenarios 1 and
2 were not included. Scenario 2 would
certainly result in lower speeds than
scenario 1. Furthermore, scenario 1 was
expected to have lower speeds than
Scenario 3. So scenarios 3 and 4 were
chosen as reasonable and representative
conditions to provide quantitative
estimates of the upper limits of hand
approach speeds.
boundaries, a series to experiments was
Figure 15: Different components of approach velocity
30 Since the testing was conducted at the UL research laboratory in Northbrook, Illinois, the test subjects were selected from the pool of employees at the campus. Some
effort was made to select candidates across a range of physical attributes and age.
page 22
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Table Saws
For these experiments, two commercially available table saws were purchased from
a national hardware retailer. One table saw was representative of "contractor" type
table saws (Figure 16). This table saw was belt-driven, incorporated integral legs and
once assembled is not typically moved around. The motor label listed the following
specifications: 120/240 V, 14/7 A, 3450 rpm (full load) and 1.75 HP. For this phase of the
project, this table saw was used during the reach-over experiments. Though the table
saw was not operational during the experiments described in this section, using an
actual table saw helped provide proper dimensioning for the experimental setup.
The other table saw was a direct-drive design, was relatively light in weight and,
although it had attached legs, could be considered a portable table saw (Figure 17).
The motor label listed the following specifications: 120 V, 60 Hz, 15 A and 5,000 rpm
(no load). This table saw was used to run the kickback experiments.
Figure 16: Photograph of belt-driven table saw used in experiments
Figure 17: Photograph of direct-drive table saw used in experiments
page 23
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Experimental Setup
Figure 18 is a photograph of the experimental setup. As the key variable to be measured
was velocity, a background grid was designed to aid in measurements. The background grid
consisted of 1 inch x 1 inch (25.4 mm x 25.4 mm) grids covering a vertical board set against
the table-top edge. Color marks at select grid points were generated to account forright- and
left-side side testing (dominant versus nondominant hand effects). Participants would wore
gloves with color marking to help track hand movement. Video of each experimental setup
for every participant was recorded using a high-speed Olympus model i-SPEED 3 camera.
The frame rate was set to 2,000 frames/s for the reach-over experiments and to 5,000
frames/s for the kickback experiments.
Figure 19 shows how the video was processed by selecting frames that covered
approximately 2 inches (51 mm) of hand displacement centered on the appropriate colored
grids. In this picture, the test subject is moving his left hand from the front side of the table
to the back side (based on cutting action). Points are selected on the hand over the specified
distance, and with the measured time duration, velocity of the hand over that portion of the
travel path is calculated.
Figure 18: Experimental setup
Figure 19: View on high-speed camera screen used to measure velocities
page 24
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Reach-over Experiments
Description: For the simulation of
reach-over conditions, an experimental
For this experimental run, a total of 12
were clearly attempting to go as fast as
people were tested with the following
possible though no such instruction was
characteristics:
procedure was designed to measure the
a.
Three women and nine men
speed of a person’s hand when reaching
b.
Ages ranging from 31 to 60 years
c.
Heights ranging from 5’4” to 6’4”
d.
Arm spans ranging from 64.5 to 78
e.
Dominant hand: 14 right-handed
across the table saw and grabbing
a stationary piece of wood with no
intervening obstacle (Figure 20). The tests
were set up to allow for reach-over with
both the left and right hand, regardless
of hand dominance. The test subject
would stand at one edge of the table
and place both hands on the edge. Upon
a signal from the experimenter, the
test subject would reach over and grab
the stationary piece of wood. For each
subject, the test was carried out with the
person's left and right hand. The intent of
these experiments was to quantify hand
velocities that could be seen if a person
had to suddenly reach for an item without
attention to any other objects.
f.
(1.63-1.93m)
inches (1.64-1.98 m)
and one left-handed
Experience ranging from "beginner"
to "intermediate" to "expert" based
on self-assessment
Results: For each test subject and each
hand, the experiment was run three
times. As participants were instructed
to simply reach over and grab the piece
of wood over a distance of 36 inches
given. From these experiments, the overall
average velocity for the 12 participants
was 12.7 ft/s (3.87 m/s) for position A
(leading edge of the blade) and 13.8 ft/s
(4.21 m/s) for position B (middle of the
blade).31 This suggests that the hand was
still accelerating as it passed the leading
edge of the blade. The maximum speed
was 22.9 ft/s (6.98 m/s) for position A
and 24.9 ft/s (7.59 m/s) for position B.
The minimum speed was 8.2 ft/s (2.5 m/s)
for position A and 8.5 ft/s (2.59 m/s) for
position B. Though the minimum speeds
cited were for two different test subjects,
the maximum speed for both positions
was achieved from the same test subject.
Table 4 and Table 5 show individual
measurements calculated for the 12
test subjects.
(0.91 m), it was noted that some subjects
Figure 20: Photograph of one test subject during reach-over experiment
31 The red-colored dots were used for tracking the reach-over using the right hand and the blue-colored dots were used for tracking the reach-over using the left hand.
page 25
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Table 4
Subject Number
1
2
3
4
5
6
page 26
LOCATION A
Hand
R
Distance (in)
2.03
1.91
1.98
Rate (ft/s)
10.24
9.96
10.33
L
2.02
2.01
1.91
10.88
11.55
10.95
R
2.18
1.98
2.07
10.66
12.24
11.87
L
2.03
1.94
1.99
12.08
11.99
10.68
R
2.07
1.95
2.06
8.85
12.47
14.96
L
2.01
2.02
2.10
11.24
12.97
13.49
R
2.11
2.00
2.11
13.03
11.92
13.04
L
2.07
1.95
2.03
12.79
14.79
14.73
R
1.99
2.07
2.07
10.05
13.82
11.51
L
2.02
2.02
2.06
12.49
9.88
9.56
R
1.96
2.01
2.06
15.53
16.74
22.93
L
2.02
1.98
1.98
13.49
14.97
18.36
Avg. Rate (ft/s)
10.18
11.13
11.59
11.58
12.09
12.57
12.66
14.10
11.79
10.64
18.40
15.61
Distance (in)
2.07
2.03
2.03
LOCATION B
Rate (ft/s)
10.78
10.56
12.06
2.04
2.06
2.02
11.71
13.76
12.05
2.02
2.06
2.12
11.24
12.24
12.64
2.11
2.07
2.02
13.03
11.15
10.22
2.08
2.06
2.06
8.47
13.76
14.95
2.04
2.08
2.03
11.74
13.32
13.01
1.97
2.11
2.11
13.13
13.50
14.65
1.98
1.94
2.06
13.78
16.18
17.20
2.11
2.06
2.11
12.53
16.31
14.64
2.10
1.97
2.07
14.55
10.59
10.47
2.02
2.06
1.94
16.86
19.11
24.91
2.00
2.07
2.11
15.17
17.23
20.72
Avg. Rate (ft/s)
11.13
12.51
12.04
11.47
12.39
12.69
13.76
15.72
14.49
11.87
20.29
17.71
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Table 5
Subject Number
7
8
9
10
11
12
page 27
LOCATION A
Hand
R
Distance (in)
1.97
2.05
1.97
Rate (ft/s)
12.64
14.87
14.28
L
1.96
2.12
2.05
12.55
13.11
15.54
R
1.98
2.03
2.03
8.26
10.90
12.10
L
1.96
2.06
2.07
9.91
12.28
12.80
R
2.20
2.04
1.90
12.70
9.72
12.69
L
2.01
1.99
2.01
10.16
12.28
9.05
R
2.01
2.04
1.97
10.82
10.95
12.19
L
1.99
1.98
2.03
9.76
10.01
10.58
R
2.00
1.97
1.94
11.10
14.31
12.94
L
1.97
1.97
2.10
11.34
15.66
17.50
R
2.04
2.02
2.09
13.10
17.72
20.50
L
2.00
1.98
2.03
15.16
14.36
16.08
Avg. Rate (ft/s)
13.93
13.73
10.42
11.66
11.70
10.50
11.32
10.12
12.78
14.83
17.11
Distance (in)
2.12
2.03
2.07
LOCATION B
Rate (ft/s)
16.08
16.88
19.15
2.09
2.18
2.15
14.53
18.15
18.87
2.02
2.02
2.06
8.65
11.24
12.74
2.02
1.99
2.02
9.92
12.28
12.97
2.02
2.07
2.15
13.49
11.10
14.96
2.13
1.98
2.02
10.75
12.25
10.22
2.06
2.05
2.09
11.08
10.69
12.91
2.03
1.97
2.04
9.93
10.60
10.95
1.99
2.24
2.11
11.41
17.79
14.67
2.11
2.03
2.09
12.14
14.07
15.82
2.02
1.99
2.06
14.65
18.44
20.21
2.03
2.14
2.03
16.88
17.00
17.77
Avg. Rate (ft/s)
17.37
17.18
10.88
11.72
13.18
11.07
11.56
10.49
14.62
14.01
17.77
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Redesigned Reach-over Experiments
Procedure: In order to create an
experimental condition closer to what
was described in the narratives, a
variation on the previous reach-over test
was developed. For these experimental
runs, the test subject was asked to push a
piece of wood (representing a work-piece)
in parallel to a surrogate blade (Figure
21) with both hands. This work-piece, in
turn, pressed against another separate
piece of board (representing the cut
work-piece) which would at some point
begin to tip over the end of the table
top. The test subject was instructed to
push the work-piece, mimicking the
The design of the test mimics several of
the narratives found in the CPSC survey:
"Near the end of the cut when the wood
started to fall. Victim reached for the
wood and put his left hand into the blade
and cut 2 of his fingers"and "Completed
a cut when wood started to fall. Victim
reached with his left hand." For each test
subject, two trials were conducted. The
first case was as described previously
while, in the second case, the piece of
board representing the cut work-piece
had a weighted end. The purpose of the
weighted end was to minimize the effect
of the operator knowing when a board
would fall.
action of an operator cutting a board and,
Results: Though this test was only run on a
upon noticing the second piece of board
small subset of the original group for the
tipping over, to reach over and grab it.
first reach-over tests, the results revealed
Discussion
The first obvious shortcoming is the small
population of test subjects. This is a common
constraint for most formalized studies
requiring human subjects. Of course, delving
into the details of human reaction, especially
for accident reconstruction, is a complex
and situationally dependent phenomenon
and is beyond the scope of this research. The
literature review showed a large and varied
body of research on measuring human reaction
times and the different influencers. From the
literature, the tests in this report would be
considered simple reaction time experiments,
as there is only a single (visual) stimulus and
a single response. The effect of influencers
on reaction time, such as fatigue, age and
drug use (Kosinski, 2012), was not studied. In
the redesigned reach-over experiments, the
number of runs per subject was minimized as
studies have shown that when subjects are
new to a reaction time task, their reaction
Figure 21: Reach-over test with falling board (left-most piece)
times show greater variability and are slower.
With practice, reaction times become faster.
For table saw injury incidents, one can assume
that the operator has not had much practice
in the action that resulted in the injury ("near
misses"). However, some instructions were
provided to the subject, and this would be
considered a "warning of impending stimuli."
Research on this topic seems to indicate that
reaction times are faster when the test subject
has been warned that a stimulus will arrive
(Jakobs et al., 2009).
page 28
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
that for three out of four test subjects, the measured speeds were slower than for the
first reach-over test procedure (Table 6). For subject 2, the measured speeds actually
increased. The average speed for this smaller group was approximately 10 ft/s (3.05 m/s).
Another tracked variable was the minimum distance of the hand from the blade as the
hand moved toward the end piece. For these runs, this distance ranged from 2 to 8 inches
(0.05–0.2 m). In none of the cases were the test subjects able to catch the falling piece.
Discussion continued
It is difficult for a subject to maintain attention
and muscle tension at a high level for more
than a few seconds. Some research focuses
on the concept of speed-accuracy trade-off
(Fairbrother, 2010). This concept simply
describes what is understood intuitively: at
higher speeds you are more likely to miss your
Table 632
target, especially under time constraints. For
table saw incidents, an operator is certainly
Location A
Location B
Subject Number
Hand Rate (ft/s)
Hand Rate (ft/s)
0
5.7
6.5
hand/finger in contact with the blade before
0
8.6
9.5
reaching the work-piece. If the person did not
11
10.4
10.7
11
11.2
11.8
5
10.2
12.1
5
10.6
9.8
2
13.3
13.7
trying to react quickly and so may be missing
the target (falling work-piece) and, in the
process, is also taking a path that brings the
feel an urgency to reach out quickly, then it
is less likely that the operator response (in a
favorable cognitive state) would lead to injury.
This entire discussion still does not inform
about how much error there might be in the
estimates of approach speeds. To consider
the confounding influence of these many
factors would be impractical and perhaps
provide only incremental insights. The
simplicity and design of these (redesigned
reach-over) experiments satisfy the intent of
having a reasonable first-order estimate for
approach speeds during reach-over related to
table saw injury incidents.
32 Subject 0 data was not shown in Table 1 (and 2) but measured an average speed of 9.8 ft/s for position A and
10.98 ft/s for position B.
page 29
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Kickback Simulations
The kickback condition is believed to be
performance requirements recommended
for table saws.
the most severe in terms of work-piece
Procedure: For the kickback simulations,
reaction of an operator to sudden kickback
"expert" was selected to run the tests.
speed. However, it is not clear how the
would alter the motion of the hand
relative to the work-piece. The worst case
is that the hand simply remains fully
coupled to the work-piece, as if it were
glued, as the work-piece draws the hand/
finger into the blade. This scenario was
deemed to be very unlikely. The other
extreme is that the work-piece moves
without accelerating the hand (full slip
condition). This is also very unlikely. As
mentioned previously, for safety reasons,
the simulation of kickback did not involve
an operator’s hand directly guiding a
work-piece and required some additional
panels to absorb any projectile motion of
the work-piece (Figure 22). Despite this
shortcoming, these experiments were
intended to provide some idea of the
speed of the work-piece during kickback
and of a possible upper limit on any
a single operator who was identified as
Two potential scenarios were explored.
The first condition mimicked closing of
the kerf slot at the completed side of
the board during a rip-cut operation. The
second condition represented the binding
of a board between a rotating blade and a
stationary guide fence during a cross-cut
operation. The portable direct-drive table
saw was used for these experiments. The
saw was equipped with a 36-count carbide
Two methods were followed to generate
contact between the board and the blade
in a manner to induce kickback. The first
method employed a short pulse force at
the edge of the work-piece to generate
board/blade contact. The second method
involved exerting a force, to close the kerf
slot, through an elastic band wrapped at
the end of the board. A wedge was inserted
into the slot initially to prevent contact.
Once the board was positioned over the
blade and the motor was powered with the
blade running at full speed, only then was
the wedge quickly removed, allowing the
board to bind to the blade.
tooth general purpose blade. To avoid any
For the cross-cut experiments, a softwood
be protected from a flying board. In all
of the edge of the board with two angled
injuries, the operator was positioned to
experiments the blade was in the fully
raised 0 degree bevel position.
For the rip-cut experiments, a pine board
measuring 1 x 2 x 19 inches (0.02 x 0.05 x
0.48 m) was precut with a kerf to extend
beyond the blade contact region.
Figure 22: Experimental setup for kickback simulations
board was prepared by removing a portion
cuts at the corners (Figure 23). The board
was then run through a cross-cut to form
two pieces. The cut board remained in
position, eventually binding to the blade.
Results: The first method for the rip-cut
experiments did not generate a kickback
condition. The second method did produce
kickback with speeds ranging from 17 to
27 ft/s (5.18 to 8.22 m/s). Calculating the
speed of the work-piece during kickback
was challenging. The motion of the
work-piece was not a simple in-plane
motion but rather a combined in-plane and
out-of-plane motion. To properly34 identify
the different components of velocity,
three points on the work-piece were used.
For the cross-cut experiments (Figure 24)
the calculated work-piece speeds for the
different components ranged from 26 to 80
ft/s (7.92 to 24.4 m/s).
33 In-plane refers to the two-dimensional plane that is the camera image.
page 30
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Figure 23: Cross-cut experiment with work-piece
Discussion
In these experiments, no approach speeds
were directly measured and so the only value
that this data can provide is an upper limit to
any discussions on approach speeds. Clearly,
a higher value of approach speed will only
increase the safety buffer, but this may come
at a cost to public safety, with few or no
technologies being able to comply with such
approach speed requirements. For now, more
research, possibly modeling,34 could still be
conducted to estimate hand approach speeds
during kickback.
Figure 24: Kickback during cross-cut experiment at different time frames starting
from the left-most image.
34 In early 2012, UL approached the Biosciences Group at the University of Michigan Transportation Research Institute for an opinion about using numerical modeling tools. They submitted a proposal that suggests that
quantitative predictions of hand motions during kickback, as an outgrowth of modeling work in vehicle crash
dummy modeling, using explicit nonlinear finite element methods, is entirely possible.
page 31
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Disclaimer
This report was prepared for research purposes only, by UL staff on the Table Saw
Working Group with review by other members of the Working Group and the Review
Panel. The information contained herein relates only to the products tested for the
purposes of this report. Neither UL LLC nor the Working Group warrants that this
information is complete or accurate or is applicable to products other than those
actually tested. This report does not mean that any product referenced herein is Listed,
Classified, Recognized or otherwise certified by UL, nor does it authorize the use of any
UL certification marks or the UL name or logo in connection with the product or system.
In no event shall UL LLC, its affiliates or any other member of the Working Group or
Review Panel or their respective organizations be liable for any damages, loss, claims,
costs or expenses arising out of or resulting from the reliance on, use or inability to use
the information contained herein. Nor shall UL LLC, its affiliates or any other members
of the Working Group or Review Panel be liable for any errors or omissions in the report.
The opinions contained herein are those of the authors and do not represent the
opinions of the external group members or the organizations that they represent.
Table Saw Safety Project
Working Group
Robert Backstrom (UL)
Paul Courtney (UL)
Peter Domeny (Power Tool Institute)
Stephen Gass, PhD (SawStop LLC)
Ted Gogoll (Power Tool Institute)
Caroleene Paul (Consumer Product
Safety Commission)
Tim Smith (Consumer Product Safety
Commission)
John Stimitz (UL)
Mahmood Tabaddor, PhD (UL)
Acknowledgments
Special Contributor
UL would like to thank the members (and their organizations) of the Working Group
Kevin Chung, MD, MS, professor of
who have volunteered their invaluable time and expertise to help guide the work
described herein. Please note that the content of this report was developed by the
authors and does not necessarily express the views and opinions of other Working
surgery, Department of Surgery,
University of Michigan at Ann Arbor
Group members, Review Panel members or their respective organizations.
Table Saw Safety Project
©2014 UL LLC All rights reserved. This paper may not be copied without permission. It is provided
for general information purposes only and is not intended to convey legal or other professional advice.
Review Panel
Don Talka, senior vice president and
chief engineer, UL
Mark Kumagai, director, division of
mechanical engineering, Consumer Product
Safety Commission
Thomas Siwek, board member,
Power Tool Institute
R. David Pittle, PhD, retired Consumer Product
Safety Commission commissioner and retired
technical director of Consumers Union
page 32
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Works Cited
Al-Qattan, M. (2012). Saw injuries causing phalangeal neck fractures in adults. Annals of Plastic Surgery, 38-40.
Beavis, R., et al. (2006). Hand trauma in workshop. Journal of Pediatric Orthopedics, 36-38.
Becker, T., et al. (1996). Tool-related injuries among amateur and professional woodworkers.
Journal of Occupational Environmental Medicine, 1032-1035.
Chowdhury, S., & Paul, C. (2011). Survey of Injuries Involving Stationary Saws:
Table and Bench Saws 2007-2008. US CPSC.
Conn, J., et al. (2005). Non-work related finger amputations in the United States: 2001-2002.
Annals of Emergency Medicine, 630-635.
CPSC. (2011). Table Saw Blade Contact Injuries; Advance Notice of Proposed Rulemaking;
Request for Comments and Information. Federal Register/ Vol. 76, No. 6.
Fairbrother, J. (2010). Fundamentals of Motor Behavior.
Hoxie, S., et al. (2009). The economic impact of electric saw injuries to the hand.
American Society for Surgery of the Hand.
Jakobs, O., et al. (2009). Effects of timing and movement uncertainty implicate the
temporoparietal junction in the prediction of the forthcoming motor actions. NeuroImage.
Justis, E. (1987). Woodworking injuries: an epidemiological survey of injuries sustained using woodworking
machinery and hand tools. Journal of Hand Surgery, 890-895.
Knight, S., et al. (2000). Injuries sustained by students in shop class. Pediatrics, 10-13.
Kosinski, R. (2012). A Literature Review on Reaction Time. Retrieved from Clemson University:
http://biae.clemson.edu/bpc/bp/lab/110/reaction.html
Landen, D., et al. (1995). Effect of recall on reporting of at-work injuries. Public Health Reports.
McGonegal, S. (2012). Table Saw Operator Blade Contact Injuries: Review and Analysis of Injury
and Societal Cost Estimates. Econometrica, Inc.
NIOSH. (1987). Injuries and Amputations Resulting from Work with Mechanical Power Presses.
Current Intelligence Bulletin 49.
Shields, B., et al. (2010). Nonoccupational table saw related injuries treated in us emergency departments,
1990-2007. The Journal of Trauma Injury, Infection and Critical Care.
Sorock, G., et al. (2001). Epidemiology of occupational acute traumatic hand injuries:
a literature review. Safety Science, 241-256.
Vesely, W. (2002). Fault Tree Handbook with Aerospace Applications. NASA Langley Research Center.
Waller, J. (1990). Disability, direct cost, and payment issues in injuries involving woodworking
and wood-related construction. Accident Analysis and Prevention, 351-360.
page 33
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Appendix A
This section shows the type of data contained in the customer injury database provided by SawStop. This is only a partial listing
of the full database that was provided to UL, which contained 1,316 entries.
Saves
1
1
1
1
1
1
1
1
1
1
1
1
1
Finger Save Recount or
Date of
Supporting Document
Accident or
Number
Date of Report
2340
3/15/2005
2341-42
3/22/2005
2343
4/23/2005
2360-61
5/5/2005
2344
5/9/2005
2332, 2345-46
5/10/2005
2348
6/13/2005
2349
6/24/2005
2347
7/22/2005
2351
8/19/2005
2336-39
8/24/2005
2350
9/2/2005
2352-53
9/28/2005
Name of Person
Customer Phone
Filling out Form
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
Time of
Accident
Name of Person
Who Touched the
Blade
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
morning REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
REDACTED
Saw Serial
No.
04480033
04480042
05060076
05110368
05020161
04450020
05150459
05210640
Saw Type
CB
CB
CB
CB
CB
CB
CB
CB
CB
CB
CB
CB
CB
Cartridge User's Years
Data Doc.
of
Fingernail Visible
No.
Experience
Body Part Contacted
Contact
Injury
Treatment
C427
Unspecified Index Finger
Yes
None
Right Pinky Finger
Yes
Bandaid
C418
Unspecified Multiple Fingers
Yes
Not Specified
C438
Not Specified
Not SpecifiedNot Specified
Right Thumb
Yes
Not Specified
Right Index Finger
Yes
Not Specified
C420
Left Thumb
Yes
Not Specified
C419
10 Right Thumb
YES
Yes
Not Specified
Not SpecifiedNot Specified
C421
Not Specified
C423
Unspecified Middle Finger
Yes
Not Specified
C422
Unspecified Pinky Finger
Yes
Not Specified
Not Specified
No
Not Specified
C424
Not Specified
Not SpecifiedNot Specified
page 34
Customer
State
AR
NC
CT
FL
WI
AZ
CO
IL
GA
CA
WA
CA
WA
Customer Name
Saw Model
31230
53230
31230
31230
51230
53230
51230
53230
51230
53480
53230
Cartridge Serial No.
not stored
0092-3304-A09
not stored
0306-0705-B01
0084-3704-A09
0124-0705-B01
0334-0805-B01
0169-1805-B02
0043-1905-B02
No.
Stitches Type of Material
Solid Wood
Plywood
Solid Wood
Not Specified
Not Specified
Solid Wood
Not Specified
Solid Wood
Not Specified
Not Specified
Solid Wood
Not Specified
Not Specified
Customer
Type
Manufacturing
Manufacturing
Education
Manufacturing
Education
Individuals
Education
Manufacturing
Individuals
Manufacturing
Manufacturing
Manufacturing
Manufacturing
Cartridge
Data
Yes
No
Yes
Yes
No
No
Yes
Yes
Yes
Yes
Yes
No
Yes
Operation
Not Specified
Rip / Straight
Not Specified
Trim Cut
Cross Cut
Dado
Rip / Straight
Rip / Straight
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Primary Safety Devices
in Use
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
None
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
In./Sec.
Type of Contact
Inadvertent Touch
Not Specified
Removing Material
Not Specified
Kickback
Kickback
Inadvertent Touch
Material shifted
Not Specified
Not Specified
Removing Material
Removing Material
Not Specified
Blade RPM
4000
4000
4000
Coast 0.8
Motor-On
Coast Down
Time (sec)
(sec.)
135
13
102
4000
4000
136
Coast 2.8
Coast 9.8
No. Teeth
Blade Type on Blade
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
8" Dado
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Estimate of Injury w/o
SawStop
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Does Not Know
Not Specified
30
2659
96
45
316
4000
page 35
Secondary Safety
Devices in Use
None/Not Specified
None/Not Specified
None/Not Specified
None/Not Specified
Miter Gauge
Push Block
Push Stick
None/Not Specified
None/Not Specified
None/Not Specified
None/Not Specified
None/Not Specified
None/Not Specified
Dollar Value
Wearing
Gloves
Kickback Feed Speed
Not SpecifiedNot SpecifiedNot Specified
Not SpecifiedNot SpecifiedNot Specified
Not SpecifiedNot SpecifiedNot Specified
Not SpecifiedNot SpecifiedNot Specified
Not Specified
Yes
Not Specified
Not Specified
Yes
Not Specified
Not Specified
No
Not Specified
Not Specified
No
Not Specified
Not SpecifiedNot SpecifiedNot Specified
Not SpecifiedNot SpecifiedNot Specified
Not SpecifiedNot SpecifiedNot Specified
Not Specified
No
Not Specified
Not SpecifiedNot SpecifiedNot Specified
Testimonial
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Yes
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Not Specified
Can SawStop
Use the
Time Between Contact
Testimonial
Contact Type
and Detection (ms)
Not Specified
50 Teeth
Not Specified
Not Specified
3 Teeth
Not Specified
16 Teeth
Not Specified
Yes
Not Specified
1 Teeth
Not Specified
1 Teeth
Not Specified
4 Teeth
Not Specified
1 Teeth
Not Specified
3 Teeth
Not Specified
Not Specified
4 Teeth
Blade Loading
Fire Type
103 Short Sum
100 Short Sum
100 Short Sum
0.8
2.8
9.8
100 Short Sum
101 Short Sum
100 Fall Off
102 Short Sum
100 Short Sum
101 Fall Off
Data is Consistent or
Inconsistent with
Contact
Consistent
No Data
Consistent
Consistent
No Data
No Data
Consistent
Consistent
Consistent
Consistent
Consistent
No Data
Consistent
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
Appendix B
This section lists some examples from the 862 narratives from the CPSC survey (Chowdhury & Paul, 2011). These narratives were
obtained by UL via FOIA request and were received in August 2011. The information in the column on the left is the case number and
the date of the accident (YYMMDD). The narrative summary in the right column is either full or an excerpts. Highlights were created
by the authors of the report.
NUMBER
DATE
NARRATIVE SUMMARY
I0850027A
080427
…after the board kicked back and his left hand was dragged across the blade while
he was using the saw.
N07C0005A
071114
14-year-old male had the tips of 3 fingers severed after he slipped on sawdust, falling
and putting his fingers in the way of a table saw that was running. He was hospitalized.
081121
54-year-old man was using a table saw and the guard’s design presents a greater danger.
He was using the saw, the piece kicked and his hand made contact with the blade,
severing his middle finger.
I0910536A
100525CCC3695
I1054000A
A 40-year-old male victim was injured while using a table saw in garage. As the victim was
ripping through an 8-foot long deck board, the board bound up and kicked back. Immediately
the blade guard assembly detached from the table saw and struck the consumer in the face.
The victim received minor lacerations and bruising to his face.
100517
081203HEP6401
081122
The 54-year-old male sustained a partial amputation of his left middle finger. The victim was
using his son’s table saw when the wood hit a knot and kicked back. This pushed his hand into
the blade. He cut and partially amputated his left middle finger.
060131HEP9001
060115
50-year-old male was using a table saw and completed a cut. When the victim reached for the
wood, his left glove got caught in the blade and he amputated the tip of his left thumb.
060214HEP9011
060202
60-year-old male. The wood jumped as he was cutting and his left hand was pulled toward
the blade. Victim cut 4 of his fingers….
060223HEP9008
060212
34-year-old male. Victim completed a cut and reached with his right hand for the wood.
Victim was not concentrating and put his right thumb into the blade and amputated the tip.
060308HEP9005
060225
32-year-old male. Cut a small piece of wood. While cutting, the victim hit a defect in the wood,
causing kickback, and the victim amputated the tip of his left middle finger.
page 36
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
060308HEP9004
060218
77-year-old male. The wood jumped while he was cutting it, and when he reached for the wood,
he cut 3 of his right-hand fingers, which included the amputation of the tip of his thumb.
060302HEP9005
060219
60-year-old male. Was on his last cut when he hit a knot in the wood. His left hand got
pulled toward the blade and he amputated his left thumb and index finger.
060323HEP9008
060104
67-year-old male completed a cut and reached for the wood and put his left thumb
into the blade.
060517HEP9001
060501
73-year-old male. Reached for a piece of wood and put his left index finger into the blade,
amputating his finger at the 1st joint.
060527HEP9003
060505
38-year-old male…and as he was cutting the stock kicked back causing the victim to
amputate his left thumb and cut his index finger.
060526HEP9001
060519
68-year-old male. Using table saw to cut a 2x4. Victim was just about through the cut when
he reached to pull the wood through and amputated his left index finger.
060526HEP9004
060509
29-year-old male. Using table saw when the wood kicked back and his left hand slammed
into the blade causing him to amputate his little finger and tip of his 4th finger.
060615HEP9003
060520
53-year-old female. Cutting with table saw and using the push stick to guide the stock.
Somehow the push stick flipped causing the stock to shift and pulling her left hand fingers into
the blade. Victim amputated 2 fingers and cut the other 3.
060710HEP9015
060605
66-year-old male. Ripping cedar. The wood kicked back causing the victim to amputate part of
his left index finger and cut the middle finger.
060728HEP9002
060707
16-year-old male. Near the end of the cut when the wood started to fall. Victim reached for
the wood and put his left hand into the blade and cut 2 of his fingers.
060720HEP9004
060707
43-year-old male. Near the end of the cut when the saw sucked the wood and victim cut his
right hand and 3 fingers.
060810HEP9001
060725
58-year-old male. Near the end of a cut, victim hit a knot in the wood as he was reaching with
his left hand to get the wood and amputated his left index finger.
060810HEP9002
060729
68-year-old male. Ripping wood when the wood kicked back and hit his hand, causing his left
thumb to go into the blade, and amputating it and part of his index finger.
page 37
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
060908HEP9001
060812
21-year-old male. Hit a knot in the wood causing the wood to slide with his left hand.
Partially amputated 3 fingers.
060710HEP9004
060611
58-year-old female. The stock kicked back causing the victim to amputate the tip of her
left index finger.
060907HEP9003
060815
56-year-old male. While cutting hit a knot in the wood which caused him to amputate the tips
of 3 of his left hand fingers.
060908HEP9006
060813
44-year-old male. Completed a cut when the wood started to fall. Victim reached with his left
hand to grab the wood and amputated his 2nd finger and cut 2 others.
061023HEP9002
061012
54-year-old male. Completed the cut. He reached for the piece of wood and put his right thumb
into the blade.
061005HEP9001
060827
68-year-old male. 1st cut when the stock kicked back and he amputated his 5th finger of
his left hand.
061124HEP9008
061028
73-year-old male. Turned off saw and reached for the wood and put his left finger into the blade.
061122HEP9002
061027
51-year-old male. Completed his first cut. Victim reached for the wood and amputated 2 left
fingers and badly lacerated 2 other fingers.
061025HEP9036
061013
60-year-male. End of 1st cut when the stock kicked back causing the victim to amputate the tip
of his left thumb.
061124HEP9003
061031
62-year-male. Near end of cut when part of frame became unglued causing the wood to slip.
Amputated the tip of his middle finger.
061128HEP9002
061029
40-year-old male. Cutting for 15 minutes and was in the middle of a cut when he hit a knot
and reached for the wood, and put his middle finger into the blade, amputating the tip.
070125HEP5121
070117
83-year-old male. Hit a knot in the wood and the stock kicked back. The stock hit him,
causing his left hand to go toward the blade and he cut the tip of his left thumb.
070125HEP9036
070118
49-year-old male. Using friend’s table saw for first time. As he was cutting, the stock kicked back
and he cut his thumb.
070126HEP0721
070124
64-year-old male. Stock kicked back causing the victim to cut his left-hand finger.
page 38
Table Saw Hazard Study on Finger Injuries Due to Blade Contact
070201HEP9039
070116
23-year-old male. Cut hardwood floors. Victim cutting for 3-4 hours when the stock kicked back
and pulled his right hand into the blade, causing him to cut his index finger and knuckle.
070210HEP6983
070124
46-year-old male. Laceration to left thumb when he accidentally reached across the back of a
semi-portable contractor’s saw to grasp a board being cut.
070226HEP8942
070217
34-year-old male. Laceration to right index finger and middle finger. While using a table saw
at home, he hit a void in a piece of Styrofoam and it sucked his hand into the saw blade.
070318HEP5601
070313
85-year-old male. Cutting small pieces of wood, a piece of wood hit a light bulb, breaking it,
startling the victim. He hit his thumb on the saw blade.
070321HEP4081
070312
62-year-old male. Cut hardwood. Guiding the wood with his right-hand, got distracted and cut
his right thumb when it got too close to the blade.
070416HEP9092
070411
66-year-male. Ripping wood. Cutting for 10-15 minutes and not sure what happened.
He believes a piece of wood hit him in the head and then his left hand fell into the blade,
cutting and or amputating all 5 fingers.
070709HEP8057
070705
44-year-male. Laceration to thumb while cutting a board. He was cutting 2x4s when a board
kicked back causing the push board to slip out of his hand and he hit the blade.
070903HEP6001
070830
18-year-male. Cut hand on table saw when it tipped on an uneven surface while cutting
wood flooring.
070920HEP5761
070905
90-year-male. Cut his finger on a bench saw when a push stick slipped and his finger went
into the blade.
080208HEP9007
080205
41-year-male. Ripping a 2x4. Push stick got caught in the blade and pulled his 2 fingers into
the blade and cut them.
page 39